Administrative Data: Impacts on Decennial Census and Research

The Demographic Research area of the Center for Economic Studies (CES) within the U.S. Census Bureau is responsible for researching and developing innovative ways to use administrative records in decennial census and survey operations. Our team of demographers, economists, geographers, and sociologists evaluate a wide array of administrative data from other federal agencies, state governments, and third party organizations. We assess the quality and coverage of these datasets and investigate how they may be useful for the Census Bureau’s data collection and processing efforts.  In addition, we use linked census, survey, and administrative records data to conduct scholarly research and to create estimates that could not be created without linked data to better inform the American Public.

Much of the work we do hinges on the ability to link records for people across different data sources. We are able to do this because another area at the Census Bureau first uses personally identifiable information (PII), such as name, date of birth, etc., to assign anonymous unique identifiers to individuals in our census, survey, and administrative data sets. They then strip off all PII and provide an anonymized file that includes these unique identifiers to researchers like myself to investigate important research questions.

One type of analysis we often perform involves linking survey data to administrative records to see if responses to survey questions match what we find in administrative records for people found in both data sources. For example, my colleagues and I linked responses from the Current Population Survey (CPS) on Medicare coverage to Medicare enrollment data and measured the extent of survey misreporting of Medicare coverage.  In this study, we found that survey responses were mostly consistent with enrollment data but we did note a small undercount of Medicare coverage in the CPS. In another case, we linked responses by American Indians and Alaska Natives regarding Indian Health Service (IHS) coverage in the American Community Survey (ACS) to IHS Patient Registration data. With this study, we found much higher levels of discordance between survey responses and administrative records. While some of the differences we found were likely due in part to definitional differences between the data sources, our analysis also suggested true inconsistencies in reporting of Indian Health Service coverage.

We also use linked data to understand how people’s responses to decennial census and survey questions change over time. For example, we have examined responses to census and survey questions on race and Hispanic origin. In one study on American Indians and Alaska Natives, we found considerable changes in racial responses between the 2000 and 2010 censuses, and by linking individuals to their responses in ACS data we were able to evaluate the characteristics of those who changed their race and those who did not.  In another project we evaluated how people reported their Hispanic origin in the 2000 and 2010 censuses and the ACS and examined the characteristics associated with a change in response, including the impact of changes in question wording and other data collection aspects.

My current work uses linked survey and administrative records data to increase our understanding of participation in social safety net programs.  This work is part of a joint project between the Census Bureau and the U.S. Department of Agriculture’s Economic Research Service and Food and Nutrition Service, as well as multiple state partners.  States that participate in the project send us data on people that receive Supplemental Nutrition Assistance Program (SNAP), Women, Infants, and Children (WIC), as well as data on Temporary Assistance for Needy Families (TANF) benefits. We link these records to ACS data to estimate eligibility and participation in each of these programs by various demographic, socioeconomic, and household characteristics.  We send our estimates of eligibility and participation back to the states with the aim of providing data that can inform program administration. For example, if we find that a particular characteristic or geographic area is associated with high rates of eligibility for a particular program but low rates of participation, it may indicate the need for further outreach.

The team I work with recently developed a visualization displaying these estimates for a few states. The visualization allows users to examine WIC eligibility and participation rates among infants and children at the county level by various characteristics. We are currently developing a similar visualization for SNAP recipients, which will include both children and adults. This is an example of the estimates we can produce with blended data that provide the public with additional information.

Renuka Bhaskar is an OSU alumna and a senior researcher in the Center for Economic Studies at the U.S. Census Bureau. Any opinions and conclusions expressed herein are those of the author and do not represent the views of the U.S. Census Bureau.

Making Sense of Census Data Resources

In my role as Ohio State’s Geospatial Information Librarian, a lot of the work that I do is related to helping researchers – at all levels and across a wide variety of disciplines – think through how they can locate, analyze, and visualize geographic data. And a lot of the time, data products provided by the U.S. Census Bureau will be relevant for addressing the research questions that they are asking.

When we hear the word “census” in 2020, our thoughts likely turn to the decennial census, and for good reason. It is hard to overstate the importance of the 2020 Census in terms of political representation and federal funding allocation, and the ways these will impact our communities over the next decade.

But it’s also important to note that census data products cover a lot more than the decennial census. In fact, the U.S. Census Bureau conducts more than 130 different surveys and programs, including the American Community Survey (ACS), Current Population Survey (CPS), Economic Census, and Longitudinal Employer-Household Dynamics (LEHD) program, to name a few.

More recently, the U.S. Census Bureau has also been releasing a variety of interesting experimental data products, which are described as “innovative statistical products created using new data sources or methodologies that benefit users in the absence of other relevant products.” Two that garnered some attention earlier this year and that have recently gone through a second phase are the Household Pulse Survey and Small Business Pulse Survey, which provide data about the social and economic effects of the COVID-19 pandemic on American households and businesses, respectively.

As mentioned in an earlier post, data products from the U.S. Census Bureau are free and publicly available. Here are a few different ways you can access these data for research, teaching, or class assignments:

U.S. Census Bureau

A lot of census data products are directly accessible in data.census.gov, a new platform that replaced American FactFinder in early 2020. The platform features a new search interface aimed at making it easier for users to locate the data they need, with more datasets planned to be added over time. It’s also possible to browse and download data tables for various programs by topic and year. If you are unable to find the data you are looking for through either of those options, you can always go directly to the website for the specific program you are interested in to see what data access options are available (and see here for the list of all surveys and programs). TIGER data products are also publicly available for working with census data in a GIS.

data.census.gov is the U.S. Census Bureau’s new platform for facilitating data access

IPUMS

IPUMS is a great resource for accessing a number of historical and contemporary census data products not readily available elsewhere. For example, NHGIS – the National Historical Geographic Information System – provides access to summary data tables and GIS-compatible boundary files from 1790 to the present and for all levels of U.S. census geography. For those working internationally, IPUMS also recently announced the launch of IHGIS – the International Historical Geographic Information System – with data tables and GIS-compatible boundary files from population, housing, and agricultural censuses from a number of countries, with more to be added over time.

Up to this point, all of the data resources I’ve been discussing have been more focused on providing summary data, presented in aggregate at different levels of U.S. census geography. But various IPUMS products also provide access to historical and contemporary census microdata, that is, individual records containing information collected about persons or households. IPUMS USA, for example, provides access to harmonized microdata from decennial censuses from 1850 to 2010 and American Community Surveys from 2000 to the present, though geographic information for these records is limited compared to summary data. IPUMS also recently announced the release of the Multigenerational Longitudinal Panel (MLP), which links individuals’ records between censuses spanning 1900-1940, with plans to extend back to 1850 in the future.

All IPUMS data products are free and publicly available, though there is a registration process required before gaining access to these data.

IPUMS provides access to various unique historical and contemporary census data products

Licensed Resources

In addition to the public data resources described above, the University Libraries licenses several resources that provide access to census data products in a fairly user-friendly way, especially for beginners. PolicyMap and Social Explorer are two examples, both of which include interactive map viewers that facilitate some geographic exploration of the data without the need to download and import data into a GIS every time. I have worked with instructors in various departments who have incorporated one of these databases into an assignment or recommended them as data sources for student projects. One other important note about Social Explorer is that it includes data tables for the 1970, 1980, 1990, and 2000 decennial censuses normalized to the 2010 census geographies to facilitate longitudinal comparisons, with data available down to the tract level.

Social Explorer has a number of interactive map viewers for exploring census data variables

This list of census data resources is by no means exhaustive, but I hope it will be a good starting point for those looking to use census data products for research, teaching, or class assignments. Have fun exploring these resources, especially if you are new to census data or less familiar with some of the other surveys and programs conducted by the U.S. Census Bureau. And if you are having trouble finding the data you need or have other questions, you can always contact a librarian.

Joshua Sadvari

Assistant Professor, Geospatial Information Librarian

University Libraries

The Ohio State University

Why We Count: Geographers and the US Decennial Census

My first job out of graduate school, before I was even finished with my PhD, was at the US Census Bureau in Washington, DC. I joined shortly after Census 2000, following the advice of one of my advisors who had spent a very happy year’s sabbatical with the Population Distribution group and believed it would be a good place for me to start my career. He wasn’t wrong. My arrival coincided with the best part of the 10-year census cycle, avoiding the dry, quiet mid-decade years, as well as the ramping-up period that immediately precedes the count, and being tasked with only the fun part: analyzing the data (Fig. 1) and helping to tell the story of a decade of U.S. population change.

Figure 1: Census 2000 Migration Analysis

Source: Migration of the Young, Single, and College Educated: 1995 to 2000, https://www.census.gov/prod/2003pubs/censr-12.pdf

Much has since changed. Census 2000 was the last decennial census to include the long form—the sample of “fun” questions that asked a subset of the population about income, education, commuting, ancestry, migration, and so on. This meant that the post-census years were overwhelmingly rich with information that hadn’t been collected for a decade. Nowadays we have the American Community Survey (ACS) to provide more timely, ongoing information and rely on the decennial census only for the “short form” data: housing type and age, race, ethnicity, sex, and household relationship for each person living in a housing unit (see Figure 2 for Census 2010 short form).

Over the years, I’ve moved on from the Census Bureau and into more traditional academic positions. However, although it has been almost 20 years since I’ve worked directly with the census, my research still depends on census data. This is common for US-based population geographers, as well as sociologists, demographers, and other social scientists—what the census lacks in detail and frequency it makes up for in geographic detail and numbers (theoretically 100 percent of the population!). Along with the ACS, the census allows us to answer “what” and “where”. Other Census Bureau surveys, such as the Current Population Survey (CPS), can offer detail and frequency, but not the tract-level information many of us need and want.

Figure 2: First page of Census 2010 Form

Source: https://www.census.gov/history/pdf/2010questionnaire.pdf

The Census is “geography” in so many more ways than published research. True, it’s all about counting people, which on the face of it sounds like demography. Scratch beneath the surface, however, and there isn’t a single aspect of the Census that is not about place and geography. Start with the purpose of the Census: the drawing and delineation of geographies of representation. Undercounts—groups and individuals the Census fails to count—are also geographical, concentrated in particular places.

In fact geography underpins every aspect of the Census Bureau’s work—an entire division of the Decennial Census programs is devoted to Geography. This makes sense: a first step to counting people is updating and maintaining address files of housing units across the entire country. For many US-based geographers, our first encounter with the Bureau is often in search of geographical and not demographic data: TIGER/line files of roads, state or county boundaries, for example.

The combination of geographic and demographic data produces other information that we consume in our academic and personal lives. Ever hear people talking about the Columbus, Ohio, metropolitan area and wonder how decisions about metro areas are made? Working with the Census Bureau, the Office of Management and Budget (OMB) decides on metropolitan (and micropolitan) standards following public consultation—for example, what ties counties together as a metropolitan unit? These rules or “delineations” are then applied to Census Bureau data and lists and geographies of metro areas are published. These geographies are used by a wide range of academic researchers and policymakers, but also—very importantly—are used for disbursement of federal funding. So we rely on the Census Bureau not only for counting people but also drawing the boundaries of everyday political life that affect every one of us.

As I grow older, I admit I am less interested in the geographies of the census and more fascinated by the way in which our census both leads and follows, where social change is concerned. What do we measure? Where? For whom? Our census is socially constructed and I believe this is something to be proud of—every decade, different questions are asked and the ways of answering evolve. Where once enumerators made decisions about the race of respondents, now individuals self-identify. Where once race was categorical and unidimensional, now the form strives to capture the nuances of identity (more than one race, but also combinations of race). Where once the “head of household”—the person completing the Census form—was assumed to be the male breadwinner, now we simply have a “householder.” Where once same-sex households were assumed to reflect respondent error, now the form explicitly makes space for a wider variety of household types. Is this perfect? No! Many of these changes were demanded: the Census Bureau followed, not led. I do believe it reflects a society wrestling with difficult concepts, though, and I would encourage each of us in our research and private lives to help push our survey instruments to continue to evolve.

Moreover, the struggle to accurately measure mutable concepts and identities is simultaneously a statistical and social quandary—this is interesting and challenging! The 2020 form, for example, still asks for respondent “sex” and offers two responses: male and female. Many would argue that while this question may be technically correct, it does not capture current social understanding of gender identity. How future forms efficiently* capture both biological/assigned-at-birth sex as well as gender—and how these are then tabulated—will be challenging in many ways, but also worthwhile. The questions we ask say as much about who we are as the responses.

A final thing to be proud of (and know about): census data are free and accessible. A strong data infrastructure has emerged over time to facilitate and preserve this access, both at the Bureau but notably through NHGIS and IPUMS.

Happy Census 2020! It’s likely to be a historical census** and possibly not in a good way (pandemics are not good for censuses). Be counted and remember that geographers count!

* because a long, complicated form is less likely to be completed

** it will be in good company; most of the 1890 files were destroyed in a 1921 fire

 

Rachel Franklin
Professor of Geographical Analysis
Spatial Analytics and Modelling Lab: SAM@Newcastle
Centre for Urban and Regional Development Studies (CURDS)
School of Geography, Politics and Sociology
Newcastle University

Editor, Geographical Analysis